基于大数据解析的煤气化过程的关键操作参数的在线估计与操作优化

For applications in coal gasification processes, aiming at its characteristics of multi-mode, nonlinear, and hard-to-measure key indices, we will study its mechanism to identify its model structure, and take into account site data to integrate ideas and methodologies of mathematical statistics, and therefore propose a comprehensive modeling approach based on first-principles analysis and data-driven methods to perform online estimation and process monitoring of important hard-to-measure indices and key operation parameters. Based upon this, we will fully mine and integrate the hidden information behind massive historical data, extract key factors that have influences on the long-term operation of the coal gasification processes, and study models suitable for multi-mode systems and comprehensive optimization methodologies that can guide optimization and adjustment of key operation parameters, to enhance the operation reliability of coal gasification processes, avoid frequent occurrence of faults, extend steady operation periods, and meanwhile improve the efficiency of carbon conversion, thereby increasing economic profit of the company. This research will promote the development and application of state estimation and dynamic soft sensing of complex production processes and system modeling and optimization techniques. The results will improve the standard of comprehensive automation of coal gasification processes, and provide a reference for extended applications to similar production processes.

以煤气化过程为应用背景，针对系统多工况、非线性、关键指标难测量等特点，深入分析其工艺机理，辨识模型结构，利用现场数据，结合数理统计思想与方法，提出基于机理分析和数据驱动方法的综合建模方法，实现重要难测指标和关键操作参数的在线估计和过程监控。在此基础上，充分挖掘并融合海量历史数据背后的隐含信息，提炼影响煤气化系统长周期运行的关键因素，研究适用于多工况系统的模型和指导生产关键操作参数优化调整的综合优化方法，增强煤气化过程运行可靠性，避免故障频发，延长稳定运行周期，同时提高碳转化效率，进而提升企业经济效益。研究工作将推动复杂生产过程状态观测、动态软测量和系统建模与优化技术的发展与应用，研究成果将提升煤气化生产过程的综合自动化水平，并对相近生产过程上的推广应用提供借鉴作用。

本课题以煤气化过程为应用背景，开展了理论紧密联系实际的研究工作。煤气化过程是一个典型的具有多工况、非线性、关键指标难测量（无测量 ）的工业过程，通过该过程的深入剖析，研究的融合机理分析建模和实验建模的综合建模方法使大数据分析在复杂工业过程中的应用成为可能，使我们能够顺利地完成立项计划，并取得了重要的研究进展。实现了重要难测指标和关键操作参数的在线估计和过程监控；提出的适用于多工况系统的模型和故障诊断方法，可以增强煤气化过程运行可靠性，避免故障频发，延长稳定运行周期；探索了不确定性条件下的指导生产关键操作参数优化调整的综合优化方法，可以提高碳转化效率，进而提升企业经济效益。研究工作在推动复杂生产过程状态观测、动态软测量和系统建模与优化技术的发展与应用做出了一定的贡献，对提升煤气化生产过程的综合自动化水平具有应用指导作用，也对相近生产过程上的推广应用提供了借鉴作用。超额完成了申请时所提出的各项指标。